jschnasse commented 8 years ago

seems all TT* ids are affected

acka47 commented 8 years ago

These are all names of corporate bodies without GND IDs. Nonetheless, they should at least appear in the nameOfContributingCorporateBody field – which they don't do. They are not taken into account in data 2.0 as well, see e.g. http://lobid.org/resources/TT002234042

@jschnasse Would it be ok for you if we only resolved this in data 2.0? Or do you need it from the production API?

jschnasse commented 8 years ago

data 2.0 will be sufficient!

acka47 commented 8 years ago

In the current 2.0 morph we don't take into account 200-1.a at all, see lines 832-922.

acka47 commented 8 years ago

154 is the corresponding ticket where we worked on adding persons without GND IDs into the RDF.

acka47 commented 8 years ago

https://github.com/hbz/lobid-rdf-to-json/issues/17 is the corresponding ticket for persons without GND ID in API 2.0.

acka47 commented 8 years ago

As discussed offline, we will concatenate ~~GND URIs (morph variable @creatorCorporateBodyId – if existing) and~~ names of corporate bodies without GND ID (not existing in morph yet) in the same order as in source to the contributorOrder string.

An example with person URIs and a string HT018700720 (snippet):

"contributorOrder": [
   "http://d-nb.info/gnd/136371671 | http://d-nb.info/gnd/129961604 | Ottovay, Kathrin | http://d-nb.info/gnd/115325859 | http://d-nb.info/gnd/2006655-7" ]

Regarding persons, the relevant part of the current morph is in lines 1130-1134.

For TT002234205 the missing part from the source data looks like this:

<datafield tag="200" ind1="-" ind2="1">
  <subfield code="a">
Rheinland-Pfalz / Ministerium für Arbeit, Soziales, Familie und Gesundheit
  </subfield>
</datafield>

The result should look like this (snippet):

{
   "id":"http://lobid.org/resources/TT002234205#!",
   "contributorOrder":[
      "Rheinland-Pfalz / Ministerium für Arbeit, Soziales, Familie und Gesundheit"
   ]
}

fsteeg commented 8 years ago

Deployed to small test set, see:

http://lobid.org/resources-small?q=hbzId:TT002234042 http://lobid.org/resources-small?q=hbzId:TT002234205 http://lobid.org/resources-small?q=hbzId:TT002234167

acka47 commented 8 years ago

The PR implements currently implements what I wrote in https://github.com/hbz/lobid/issues/302#issuecomment-212873147 and will suffice for presentation purposes. As I wrote in https://github.com/hbz/lobid/issues/302#issuecomment-209889538, the strings are also missing in contributingCorporateBodyLabel which is used for querying – at least by NWBib. @jschnasse: Do you also search via the contributingCorporateBodyLabel? If yes, we would also have to add the string in there..

jschnasse commented 8 years ago

I use only contributorOrder from data 2.0. Testset looks fine for me.

acka47 commented 8 years ago

+1 I forgot that nobody uses API 2.0 for searching anyway, thus we don't need no further adjustments here. (I don't know how many of these corporate bodies without GND ID exist and whether we should make them searchable. If yes, we would have to do it in production.)

fsteeg commented 8 years ago

Will be deployed as part of full index build over the weekend.

fsteeg commented 8 years ago

Deployment as part of weekly full indexing failed, see https://github.com/hbz/lobid/issues/304

fsteeg commented 8 years ago

Deployed to beta, closing. See:

http://lobid.org/resources/TT002234042 http://lobid.org/resources/TT002234205 http://lobid.org/resources/TT002234167

dr0i commented 7 years ago

The contributor was only added to the contributorOrder field. There was no dc:contributor. With hbz/lobid-resources#125 we get rid of contributorOrder. A contributor must be added to the contribution-list.

dr0i commented 7 years ago

See http://lobid.org/resources/TT002234042 as an example (Lierscheid).

dr0i commented 7 years ago

Deployed to staging, see http://test.lobid.org/resources/TT002234042.

acka47 commented 7 years ago

+1

hbz / lobid

Include names of corporate bodies without GND ID #302

154 is the corresponding ticket where we worked on adding persons without GND IDs into the RDF.